MEASUREMENT OF RECURRING VERSUS NON-RECURRING CONGESTION · MEASUREMENT OF RECURRING VERSUS...
Transcript of MEASUREMENT OF RECURRING VERSUS NON-RECURRING CONGESTION · MEASUREMENT OF RECURRING VERSUS...
Final ReportResearch Project T2695, Task 36
Congestion Measurement
MEASUREMENT OFRECURRING VERSUS NON-RECURRING CONGESTION
by
Mark E. HallenbeckDirector
John M. Ishimaru Jennifer NeeSenior Research Engineer Research Engineer
Washington State Transportation Center (TRAC)University of Washington, Box 354802
University District Building1107 NE 45th Street, Suite 535
Seattle, Washington 98105-4631
Washington State Department of TransportationTechnical MonitorToby D. Rickman
State Traffic Engineer
Prepared for
Washington State Transportation CommissionDepartment of Transportation
and in cooperation withU.S. Department of Transportation
Federal Highway Administration
October 2003
TECHNICAL REPORT STANDARD TITLE PAGE1. REPORT NO. 2. GOVERNMENT ACCESSION NO. 3. RECIPIENT'S CATALOG NO.
WA-RD 568.2
4. TITLE AND SUBTITLE 5. REPORT DATE
MEASUREMENT OF RECURRING VERSUS October 2003NON-RECURRING CONGESTION 6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) 8. PERFORMING ORGANIZATION REPORT NO.
Mark E. Hallenbeck, John Ishimaru, and Jennifer Nee9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. WORK UNIT NO.
Washington State Transportation Center (TRAC)University of Washington, Box 354802 11. CONTRACT OR GRANT NO.
University District Building; 1107 NE 45th Street, Suite 535 Agreement T2695, Task 36Seattle, Washington 98105-463112. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
Research OfficeWashington State Department of TransportationTransportation Building, MS 47372
Final Report
Olympia, Washington 98504-7372 14. SPONSORING AGENCY CODE
Doug Brodin, Project Manager, 360-705-797215. SUPPLEMENTARY NOTES
This study was conducted in cooperation with the U.S. Department of Transportation, Federal HighwayAdministration.16. ABSTRACT
A research effort was initiated to develop and test a methodology to estimate the extent to which urban
congestion is produced by the routine presence of large numbers of vehicles on a facility (so-called
recurring congestion) versus unexpected disruptions or other events (non-recurring congestion),
particularly lane blocking incidents. An analysis was performed using a preliminary methodology to
estimate the magnitude, extent, and relative causes of congestion on urban freeways in the central Puget
Sound region of Washington State. This information is designed to assist transportation agencies in 1)
selecting the most appropriate strategies to improve freeway operations, and 2) allocating available
resources to achieve those improvements. This paper documents the preliminary methodology and results
from this ongoing research project.
17. KEY WORDS 18. DISTRIBUTION STATEMENT
Recurring congestion, non-recurring congestion,transient congestion, freeway performancemonitoring, ITS data archiving, incident database
No restrictions. This document is available to thepublic through the National Technical InformationService, Springfield, VA 22616
19. SECURITY CLASSIF. (of this report) 20. SECURITY CLASSIF. (of this page) 21. NO. OF PAGES 22. PRICE
None None
iii
DISCLAIMER
The contents of this report reflect the views of the authors, who are responsible
for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of the Washington State Transportation
Commission, Department of Transportation, or the Federal Highway Administration.
This report does not constitute a standard, specification, or regulation.
iv
v
CONTENTS
Section Page
Recurring Versus Non-Recurring Congestion............................................................ 1
Research Approach ................................................................................................... 4
Summary Project Findings ........................................................................................ 8
Detailed Analytical Results ..................................................................................... 12
Additional Information and Limitations of the Databases ........................................ 12
vi
TABLES
Table Page
1. Daily Delay Computations using a 50 mph Baseline Condition ..................... 12
2. Daily Delay Computations using a 60 mph Baseline Condition ..................... 13
3. Daily Delay Computations Descriptive Corridor Statistics............................. 14
4. Congestion Caused by Lane Blocking Incidents (Conservative Estimate,50 mph Baseline) by Time Period............................................................... 15
5. Congestion Caused by Lane Blocking Incidents (Liberal Estimate,50 mph Baseline) by Time Period............................................................... 16
6. Total Non-Recurring Congestion (50 mph Baseline) by Time Period............. 17
1
MEASUREMENT OFRECURRING VERSUS NON-RECURRING CONGESTION
This project examined the size, extent, and relative causes of congestion on urban
freeways in the Puget Sound region. It made a substantial (but initial) step toward
understanding the relative importance of different causes of congestion in the
metropolitan area. The intent was to provide the Washington State Department of
Transportation (WSDOT) with the knowledge it needs to make decisions about how to
improve freeway operations and how to allocate the resources available for making those
improvements.
Key to this study was an analysis of how much congestion is caused simply
because too many vehicles routinely attempt to use the facility, and how much is caused
by frequent, “unexpected” disruptions or events.
RECURRING VERSUS NON-RECURRING CONGESTION
For this project, congestion was subdivided into “recurring” and “non-recurring”
congestion. “Non-recurring” congestion was further subdivided into delay caused by
lane blocking incidents and other causes of “unusual congestion.”
Recurring congestion was defined by this study as congestion caused by routine
traffic volumes operating in a typical environment. In layman’s terms it might be thought
of as “the congestion present on a normal day if nothing bad has happened on the
roadway.” In essence, this definition is grounded in the concept of “expected
congestion” if no “unusual circumstances” occur.
While this definition makes sense, it does create some analytical difficulties.
Most important is the fact that “expected congestion” is dependent on time and location.
2
On most urban freeways, more congestion is expected on weekdays in the commute
periods than during the middle of the day. Similarly, more congestion is expected in the
middle of the day than at night. Friday mornings tend to have light congestion, while
Friday afternoon tends to be worse. Weekends tend to have lower “expected” levels of
congestion than weekdays, although most people expect very heavy congestion during
specific holiday weekends (e.g., the Sunday afternoon on Memorial Day weekend).
Similarly, expected conditions change with location. Most urban freeways have
known bottlenecks, locations where traffic backs up routinely, especially during peak
commute periods. Other roadway sections experience little routine congestion.
“Non-recurring congestion” was defined for this study as “unexpected or unusual
congestion caused by an event that was unexpected and transient relative to other similar
days.” Non-recurring congestion can be caused by a variety of factors, including, but not
limited to,
• lane blocking accidents and disabled vehicles
• other lane blocking events (e.g., debris in the roadway)
• construction lane closures
• significant roadside distractions that alter driver behavior (e.g., roadsideconstruction, electronic signs, a fire beside the freeway)
• inclement weather
• heavier than normal vehicle merging movements
• significant increases in traffic volume in comparison to “normal” trafficvolumes.
3
Under low traffic volume conditions (relative to capacity), none of these events may
result in congestion. But as volume grows relative to capacity, congestion can easily
result from these events.
Reducing the impacts of these unexpected events during moderate to high volume
conditions is one of the major goals of traffic management systems. Many of these same
traffic management efforts can also be used to control and limit recurring congestion.
The question for highway agencies is how much congestion is caused by simply having
too many vehicles routinely using the roadways (recurring congestion), and how much is
caused by short-term, unexpected events (non-recurring congestion)?
Understanding and tracking this distribution may allow the agency to
• select cost-effective approaches for lowering the mount of congestionexperienced by travelers
• track the effectiveness of the congestion relief efforts the agencyundertakes
• determine the limits of likely operational improvements, and thus identifycongestion that can be alleviated only by capacity expansion througheither mode shift or an increased number of roadway lanes.
Within the category of non-recurring congestion, it is also important for highway
agencies to understand how much congestion is caused by particular types of events, as
the causes of these “temporary” problems may change how an agency would approach
limiting their effects. This study examined only the simple split of lane blocking versus
“other” non-recurring events. This split was intended to address the need for incident
response programs aimed at quickly arriving at, and clearing, lane blockages. However,
these programs are likely to have a limited effect in the face of many other causes of non-
4
recurring congestion. Thus, additional study will be needed to identify these other causes
and their relative importance to improving traffic flow.
RESEARCH APPROACH
To simplify this project while dealing with the issues of temporal and spatial
variability in traffic flow, the research team restricted this initial analysis of recurring and
non-recurring congestion to weekdays, and specifically, Tuesday through Thursday. In
addition, the analysis was broken into four specific time periods: AM peak (6:00 to 9:00
AM), midday (9:00 AM to 3:00 PM), PM peak (3:00 to 7:00 PM), and night.
Geographically, the study included the mainlines of the entire central Puget Sound
metropolitan freeway system under surveillance. That includes five separate, connected
freeways and roughly 100 center-line miles of roadway.1 Data for two months were used,
covering September and October of 2002.
Although the times and locations were limited, the approach we have developed is
applicable to any time period and geographic scope. Only the conclusions produced in
this report are limited to these facilities and time periods.
The key to the analysis was the mathematical definition of “recurring” and “non-
recurring” congestion. We chose to define “non-recurring congestion” and to then
associate any congestion not associated with those non-recurring conditions as being
“recurring.” For this study, we used the following definitions:
“Non-recurring Congestion” is all congestion that occurs when conditions are
significantly worse than expected, that is, worse than routine operating conditions.
Recurring Congestion is all other congestion.
1 35 miles on I-5, 28 miles on I-405, 13 miles on I-90, 12 miles on SR 520, and 11 miles on SR 167.
5
Routine Operating Conditions are the median operating condition for that time
and location on the road network.
Mathematically, we used percentage of lane occupancy, aggregated at 5-minute
intervals across all (directional) lanes of a freeway at each data collection location, as our
measure of facility operating condition, along with vehicle volume and speed. We
defined “non-recurring congestion” mathematically as occurring when lane occupancy is
5 or more percentage points higher than the median operating condition (expressed in
percentage of lane occupancy) for that time and location for all days being studied that do
not contain a lane blocking incident during the period of interest. Thus, if between 6:00
and 6:05, at milepost 100, the roadway routinely operates with a median lane occupancy
of 10 percent, non-recurring congestion only exists when measured lane occupancy
exceeds 14.9 percent. A more detailed discussion of how this definition was chosen is
presented in the Technical Report for this project.2
The estimation of the amount of actual congestion at any point in time and space
is based on the volume and speed of vehicles present at each loop measurement location,
every 5 minutes. The speed of vehicles is compared to a reference speed, and vehicle-
hours of delay is computed on the basis of the difference between average speed and that
reference speed. Two different reference speeds are used, one of 60 miles per hour, the
speed limit for most of the freeway system being studied, and one of 50 mph, the
approximate speed at which the Highway Capacity Manual indicates maximum facility
flow can be maintained. The 60 mph reference can be thought of as a measure of “delay”
from the perspective of the motorist, who wishes to travel at the speed limit. The 50 mph
2 John Ishimaru, “Measurement of Recurring and Non-recurring Congestion: Technical Report,” October2003.
6
reference speed can be thought of as “delay” from the perspective of a highway agency
trying to maximize the use of existing roadway capacity.
The following list describes the steps in the delay computation procedure. All
procedures are performed by facility, by direction.
• Determine traffic performance (vehicle volume, speed, and lane
occupancy) by time and location.
• Identify which days were affected by lane blocking incidents. (This is
performed by time of day, where the time of day is defined as AM peak
period, midday, and PM peak period.)
• For all days when lane blocking incidents did not occur during the period
of analysis, compute the median condition (by time of day and location).3
This median condition serves as the “expected, recurring, condition.”
• For each day, compute the times and locations where congestion was
“significantly worse”4 than the “expected recurring condition.” These
locations were defined for this study as being sites of “non-recurring
congestion.”
• For all days when major lane blocking incidents took place, determine the
time, location, and duration of each incident recorded.
3 Note that if a lane blocking incident occurred in the AM peak period and the effect of that incident lastedinto the midday period, that midday was also treated as an “incident day.”4 “Significantly worse” is defined as a change in lane occupancy of greater than 5 percent.
7
• For all days when major lane blocking incidents took place, determine the
geographic and temporal extent of “significantly worse than expected”
congestion5 that was associated with each lane blocking incident.6
• Using the 60 mph baseline standard, compute the delay associated with
each of the “significantly worse than expected” geographic and temporal
areas associated with lane blocking incidents. (This is performed twice,
once with a “liberal” definition and once with a “conservative” definition.)
These are the estimates of non-recurring delay caused by lane blocking
incidents.
• Using the 60 mph baseline speed, compute the delay associated with all
areas where and times when conditions are defined as “significantly worse
than expected” for all days. This includes the delay associated with
incidents, as well as all other non-recurring delay. This is the estimate of
total non-recurring delay.
• For all days compute the total amount of delay, assuming delay is any
travel slower than free flow conditions (60 mph).
5 Because of the interaction of the many factors that affect traffic congestion, this task is extremelydifficult. The project team chose two alternative approaches to this task. One provided a conservativeestimate of “incident related congestion,” meaning that much of the congestion occurring in the vicinity ofthe incident after it had been cleared was attributed to background traffic volumes and was not associatedwith the incident. The second approach assigned the majority of congestion contiguous to the location ofthe incident and after its occurrence to that incident. See the Technical Report cited previously for a moredetailed discussion of how congestion was assigned to a lane blocking incident under these two differentapproaches.6 Note that one limitation of this analysis methodology is that the project team did not attempt to track theeffects of incident congestion across facilities. That is, we did not attempt to determine the extent to whichdelay on I-5 was caused by a lane blocking incident that occurred on SR 520. Thus, our estimates of laneblocking-related, non-recurring delay slightly underestimated the actual values.
8
• Subtract from the estimate of total delay the non-recurring delay computed
above. The result is the estimate of total recurring delay based on a 60
mph standard.
• Repeat the preceding four steps, except with a 50 mph baseline speed.
This produces the required estimates of recurring, non-recurring, and total
delay based on the 50 mph standard.
• Aggregate and summarize these levels of delay across corridors and for
different time of day/volume conditions within corridors.
SUMMARY PROJECT FINDINGS
Our review of the performance and incident data for these two months in 2002
provided the following insights. (Some additional notes on these conclusions follow
these bullets.)
• Corridor performance, including total delay, the percentage of delay that is
related to non-recurring sources, and the percentage of delay associated
with lane blocking incidents, is highly variable.
The primary sources of variation are vehicle volumes, especially as they
vary as they approach roadway capacity, the length of time at which
volumes routinely approach capacity, the number of incidents that occur
along the corridor, and the frequency with which those incidents block
freeway lanes.
• For the urban freeway corridors examined, lane blocking incidents are
responsible for between 2 and 20 percent of total daily delay. Facilities
9
that have few lane blocking incidents, primarily I-90 and I-405
northbound from Bellevue to Mill Creek, are on the lower range of these
figures, whereas facilities with larger numbers of lane blocking incidents
(SR 520, I-5) are on the upper end.
• Non-recurring delay generally ranges between 30 to 50 percent of all peak
period, peak direction delay, but it is between 30 and 70 percent of total
daily delay. Outliers occur where the frequency or severity of incidents is
particularly high, or where traffic volumes are very low relative to
roadway capacity.
• Lane blocking incidents generally account for between 10 and 35 percent
of all non-recurring delay.
• Thus, for most corridors, only between 1 and 10 percent of peak period
delay is caused by lane blocking incidents. In part this is due to the high
level of “recurring congestion” on the area’s freeways and the relative
infrequency of lane blocking incidents during the peak period. These
numbers can be exceeded when a large number of lane blocking incidents
occur, or when little other congestion takes place.
• For all roadways, the more that recurring congestion occurs, the lower the
percentage of “non-recurring” congestion from all sources. For example,
on a frequently congested facility such as I-405 northbound from Tukwila
to Bellevue, only a modest percentage of congestion is caused by non-
recurring events (including lane blocking events). The opposite is also
true. Late at night, when almost no recurring congestion exists, essentially
10
all delay is caused by lane blocking events. This is because without
significant reductions in available roadway capacity, there is no
congestion.
• Expressing the significance of non-recurring congestion only as a
percentage of total congestion does not fully describe the magnitude of
non-recurring congestion such as lane blocking events. For example, even
though the percentage of congestion that is non-recurring decreases as
volume on a facility increases, the absolute number of vehicle hours of
delay due to non-recurring events can increase significantly. (This
dichotomy occurs simply because of mathematical relationships of ratios.)
• In peak periods, on any facility, a lane blocking incident of even a short
duration tends to result in substantial delay. For example, a 3-minute, lane
blocking incident on eastbound I-90 that occurred at 9:00 AM (after the
peak of the rush hour) resulted in 35 vehicle-hours of delay on that
freeway (roughly 700 minutes of vehicle-delay for each minute of incident
duration).
The combination of these facts leads to the following conclusions.
• While incident response is as important as it has been made out to be, the
Department still has a long way to go in reducing the effects of “unusual”
occurrences. Many large delays still occur for which incidents are not
responsible, and for which no “cause” is present in the WITS7 database.
7 The Washington Incident Tracking System database (WITS) tracks the activities of the WSDOT freewayservice patrol.
11
(This means that there was no on-site WSDOT response to whatever
caused the “unusual” delay.)
These “extra” delays could be caused by any of a large number of factors,
including
* unusual volume surges at ramps (e.g., on I-90 westbound at West
Lake Sammamish) that are not being handled effectively by the ramp
metering program (This could in turn be caused by a lack of ramp
storage space or other control issues.)
* weather, although the period selected for study (September/October of
2002) does not have the worst weather of the year
* visual distractions (police vehicle by the side of the road) or
something occurring outside of the right-of-way.
Additional effort needs to be spent to understand the causes of these non-recurring
congestion events.
• Even in places where congestion occurs routinely, the scope and intensity
of that congestion varies considerably from day to day.
• This implies that there is still considerable room for improvement in
increasing the reliability of the freeway system through better freeway
management.
12
DETAILED ANALYTICAL RESULTS
The six tables at the end of this report summarize the basic results of this analysis.
Table 1 describes the level of daily delay experienced on each of the study corridors if a
50 mph baseline standard (optimal flow) is used as the basis for delay calculations. Table
2 describes corridor delay using the 60 mph baseline standard (free flow conditions).
Table 3 provides some basic statistics on each of the corridors, including its length, the
width of the “average” directional section for each corridor (this changes from segment to
segment within a corridor), the number of lane blocking incidents recorded by WSDOT
during the two-month study, and the resulting delay per lane-mile calculated for each
facility.
Tables 4, 5, and 6 show how delay varies across the three primary time periods
during which delay occurs (AM, midday, and PM). Table 4 shows delay related to lane
blocking incidents using the “conservative” method for associating specific delay to a
given incident. Table 5 shows how these values change if the more liberal association of
congestion is used. Table 6 shows the distribution of all “non-recurring” congestion.
ADDITIONAL INFORMATION AND LIMITATIONS OF THE DATABASES
All facility performance data used in this study were extracted from the 5-minute
freeway flow archive maintained by WSDOT. Data on the time, location, and duration of
all lane blocking incidents used in the study area were obtained by comparing two major
data sources, the Washington Incident Tracking System database (WITS), which tracks
the activities of WSDOT’s freeway service patrols, and the Traffic Systems Management
Center (TSMC) incident log. Comment fields in both databases were manually reviewed
to more precisely determine the presence, location, and duration of lane blocking
13
incidents. While this was time consuming, it was necessary to resolve discrepancies
found in both incident databases. Removing these discrepancies was necessary to ensure
the integrity of the project conclusions.
The existing CD Analyst software and analytical process were then revised to
perform the analytical steps described earlier in this paper.
By using only Tuesday through Thursday flow data during September and
October, the analysis was able to remove some known variation in traffic volumes. This
carefully selected timeframe allowed a more precise definition of “routine, non-incident,
traffic conditions,” and thus the computation of “non-recurring” congestion. However, it
also skewed the results toward Tuesday-Thursday travel conditions, and thus results in a
slightly biased conclusion. Inclusion of Monday, Friday, and weekend traffic
performance data would result in slightly different results. However, the inclusion of
these days would also complicate the analysis by making it more difficult to define
routine, non-incident conditions. The reason is that because motorists expect weekend
conditions to be different, different baseline conditions would be needed to perform this
analysis from the perspective of the traveler. It is not clear without additional research
which of these four days should be treated independently from the “average weekday”
condition.
Before extending the analysis to these additional days, the project team wanted to
present this approach to the roadway performance research community to obtain input
and comments. The project team has particular interest in the research community’s
review of this study’s use of alternative baseline speeds for the calculation of delay, the
14
use of the median condition for determining “routine” conditions, and the use of “routine
conditions” along with a modest performance band to define “non-recurring” congestion.
15
Tab
le 1
: D
aily
Del
ay C
ompu
tati
ons
Usi
ng a
50
mph
Bas
elin
e C
ondi
tion
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
-E
vere
tt
WB
SR
520
Red
mon
d-
Seat
tle
EB
SR
520
Seat
tle -
Red
mon
d
WB
I-
90 Issa
quah
- Se
attle
EB
I-9
0Se
attle
–Is
saqu
ah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-40
5Sw
amp
Cre
ek -
Bel
levu
e
SB I
-405
Bel
levu
e- T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
-R
ento
n
Tot
al D
elay
(Tue
s-T
hurs
,fo
r Se
pt/O
ct)
(veh
.-hr
s.)
161,
000
23,0
0098
,000
103,
000
29,0
0035
,000
29,0
0019
,000
72,0
0065
,000
85,0
0075
,000
40,0
0016
,000
Perc
enta
ge o
fD
elay
Cau
sed
by L
ane
Blo
ckin
gIn
cide
nts
(Con
serv
ativ
eE
stim
ate)
8.1%
18.6
%8.
8%6.
9%10
.1%
6.1%
1.8%
2.5%
3.9%
0.4%
4.1%
3.2%
4.9%
6.5%
Perc
enta
ge o
fD
elay
Cau
sed
by L
ane
Blo
ckin
gIn
cide
nts
(Lib
eral
Est
imat
e)15
.1%
19.3
%16
.6%
16.8
%23
.3%
9.4%
2.%
5.6%
4.6%
1.6%
8.1%
4.2%
12%
7.9%
Perc
enta
ge o
fA
ll D
elay
Tha
tIs
Non
-R
ecur
ring
52.2
%70
.5%
43.%
49.6
%60
.4%
35.6
%68
.2%
53.3
%37
.4%
19.5
%40
.3%
41.4
%65
.2%
77.5
%
16
Tab
le 2
: D
aily
Del
ay C
ompu
tati
ons
Usi
ng a
60
mph
Bas
elin
e C
ondi
tion
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
-E
vere
tt
WB
SR
520
Red
mon
d-
Seat
tle
EB
SR
520
Seat
tle –
Red
mon
d
WB
I-
90 Issa
quah
- Se
attle
EB
I-9
0Se
attle
–Is
saqu
ah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-405
Swam
pC
reek
-B
elle
vue
SB I
-405
Bel
levu
e- T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
-R
ento
n
Tot
al D
elay
(Tue
s-T
hurs
,fo
r Se
pt/O
ct)
(veh
.-hr
s.)
208,
000
48,0
0012
7,00
014
5,00
056
,000
48,0
0040
,000
29,0
0010
3,00
091
,000
113,
000
100,
000
55,0
0034
,000
Perc
enta
ge o
fD
elay
Cau
sed
by L
ane
Blo
ckin
gIn
cide
nts
(Con
serv
ativ
eE
stim
ate)
6.8%
9.8%
7.5%
5.4%
6.%
4.9%
1.5%
1.9%
3.%
0.3%
3.4%
2.7%
3.9%
3.4%
Perc
enta
ge o
fD
elay
Cau
sed
by L
ane
Blo
ckin
gIn
cide
nts
(Lib
eral
Est
imat
e)12
.8%
10.2
%14
.3%
13.1
%13
.6%
7.6%
1.7%
4.1%
3.6%
1.3%
6.7%
4.%
13.3
%4.
1%
Perc
enta
ge o
fA
ll D
elay
Tha
tIs
Non
-R
ecur
ring
44.9
%38
.1%
37.3
%39
.9%
36.3
%29
.1%
57.2
%42
.1%
29.8
%15
.9%
33.7
%25
.1%
51.6
%40
.1%
17
Tab
le 3
: D
aily
Del
ay C
ompu
tati
ons
Des
crip
tive
Cor
rido
r St
atis
tics
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
-E
vere
tt
WB
SR
520
Red
mon
d-
Seat
tle
EB
SR
520
Seat
tle -
Red
mon
d
WB
I-
90 Issa
quah
- Se
attle
EB
I-9
0Se
attle
–Is
saqu
ah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-405
Swam
pC
reek
-B
elle
vue
SB I
-405
Bel
levu
e- T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
-R
ento
n
Tot
al D
elay
(Tue
s-T
hurs
,fo
r Se
pt/O
ct)
(veh
.-hr
s.)
161,
000
23,0
0098
,000
103,
000
29,0
0035
,000
29,0
0019
,000
72,0
0065
,000
85,0
0075
,000
40,0
0016
,000
Len
gth
ofC
orri
dor
(mile
s)22
.512
.513
2211
.511
13.5
12.5
1315
15.5
12.5
99
Wid
th o
fA
vera
geR
oadw
aySe
ctio
n(N
umbe
r of
lane
s)3
44
32
23
32
33
22
2
Num
ber
ofL
ane
Blo
ckin
gIn
cide
nts
Rec
orde
d(S
ept –
Oct
2002
)45
1837
2728
248
828
726
1911
8
Del
ay P
erL
ane-
Mile
(veh
.-hr
s.)
7,20
01,
900
7,50
04,
700
2,60
03,
200
2,10
01,
500
5,50
04,
300
5,50
06,
000
4,40
01,
700
18
Tab
le 4
: C
onge
stio
n C
ause
d B
y L
ane
Blo
ckin
g In
cide
nts
(Con
serv
ativ
e E
stim
ate,
50
mph
Bas
elin
e) B
y T
ime
Per
iod
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
-E
vere
tt
WB
SR 5
20R
edm
ond
- Se
attle
EB
SR 5
20Se
attle
–R
edm
ond
WB
I-90
Issa
quah
- Se
attle
EB
I-90
Seat
tle –
Issa
quah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-405
Swam
pC
reek
-B
elle
vue
SB I
-405
Bel
levu
e– T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
–R
ento
nA
M P
eak
Per
iod
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
2,10
050
2,10
060
020
030
010
010
01,
100
01,
900
700
100
700
Perc
enta
geof
Tot
al A
MPe
ak P
erio
dD
elay
2%4%
5%10
%2%
3%1%
1%3%
0%3%
6%17
%8%
Num
ber
ofD
ays
With
Lan
eB
lock
ing
Inci
dent
s10
110
56
44
210
010
31
4M
idda
yT
ues-
Thu
rsV
eh.-
hrs
ofD
elay
)5,
100
1,70
03,
600
2,20
01,
100
500
100
100
1,10
00
800
300
500
200
Perc
enta
geof
Tot
alM
idda
yD
elay
17%
29%
14%
28%
25%
9%7%
5%5%
0%6%
2%10
%5%
Num
ber
ofD
ays
With
Lan
eB
lock
ing
Inci
dent
s16
916
78
83
39
09
44
4P
M P
eak
Per
iod
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
5,80
02,
600
2,90
04,
300
1,60
01,
300
300
200
600
200
800
1,40
01,
400
50Pe
rcen
tage
of T
otal
PM
Peak
Per
iod
Del
ay14
%16
%11
%5%
6%7%
3%3%
5%0%
10%
3%4%
3%N
umbe
r of
Day
s W
ithL
ane
Blo
ckin
gIn
cide
nts
198
1115
1212
33
87
713
63
19
Tab
le 5
: C
onge
stio
n C
ause
d B
y L
ane
Blo
ckin
g In
cide
nts
(Lib
eral
Est
imat
e, 5
0 m
ph b
asel
ine)
By
Tim
e P
erio
d
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
-E
vere
tt
WB
SR
520
Red
mon
d-
Seat
tle
EB
SR
520
Seat
tle -
Red
mon
d
WB
I-
90 Issa
quah
– Se
attle
EB
I-9
0Se
attle
–Is
saqu
ah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-405
Swam
pC
reek
-B
elle
vue
SB I
-405
Bel
levu
e- T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
-R
ento
nA
M P
eak
Per
iod
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
3,30
050
2,40
060
050
030
010
010
01,
300
02,
900
800
100
1,00
0Pe
rcen
tage
of T
otal
AM
Peak
Per
iod
Del
ay4%
4%5%
10%
5%3%
1%1%
3%0%
5%6%
17%
7%N
umbe
r of
Day
s W
ithL
ane
Blo
ckin
gIn
cide
nts
101
105
64
42
100
103
14
Mid
day
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
8,50
01,
700
6,50
03,
700
2,10
01,
000
200
200
1,10
00
1,90
030
050
020
0Pe
rcen
tage
of T
otal
Mid
day
Del
ay29
%29
%25
%46
%46
%18
%10
%6%
6%0%
15%
2%11
%5%
Num
ber
ofD
ays
With
Lan
eB
lock
ing
Inci
dent
s16
916
78
83
39
09
44
4P
M P
eak
Per
iod
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
12,5
002,
700
7,40
013
,000
4,20
02,
000
300
800
900
1,00
02,
100
2,20
04,
200
50Pe
rcen
tage
of T
otal
PM
Peak
Per
iod
Del
ay29
%17
%27
%15
%15
%10
%3%
8%7%
2%25
%4%
12%
3%N
umbe
r of
Day
s W
ithL
ane
Blo
ckin
gIn
cide
nts
198
1115
1212
33
87
713
63
20
Tab
le 6
: T
otal
Non
-Rec
urri
ng C
onge
stio
n (5
0 m
ph b
asel
ine)
By
Tim
e P
erio
d
SB I
-5E
vere
tt -
Seat
tle
SB I
-5Se
attle
CB
D–
Des
Moi
nes
NB
I-5
Des
Moi
nes
–Se
attle
CB
D
NB
I-5
Seat
tleC
BD
–E
vere
tt
WB
SR
520
Red
mon
d-
Seat
tle
EB
SR
520
Seat
tle –
Red
mon
d
WB
I-
90 Issa
quah
- Se
attle
EB
I-9
0Se
attle
–Is
saqu
ah
NB
I-4
05T
ukw
ila -
Bel
levu
e
NB
I-4
05B
elle
vue
– Sw
amp
Cre
ek
SB I
-405
Swam
pC
reek
-B
elle
vue
SB I
-405
Bel
levu
e- T
ukw
ila
SB S
R16
7R
ento
n -
Aub
urn
NB
SR
167
Aub
urn
-R
ento
nA
M P
eak
Per
iod
Tue
s-T
hurs
Veh
.-hr
s of
Del
ay)
33,1
0080
011
,500
3,50
02,
700
1,80
09,
900
2,80
012
,000
100
19,6
006,
600
500
7,30
0Pe
rcen
tage
of T
otal
AM
Peak
Per
iod
Del
ay37
%68
%26
%60
%35
%17
%58
%44
%30
%29
%31
%51
%92
%56
%M
idda
yT
ues-
Thu
rsV
eh.-
hrs
ofD
elay
)21
,000
3,30
017
,500
6,90
04,
000
2,80
01,
400
2,00
09,
500
900
8,50
07,
700
2,80
03,
300
Perc
enta
geof
Tot
alM
idda
yD
elay
71%
57%
68%
88%
87%
52%
86%
69%
48%
33%
64%
48%
56%
75%
PM
Pea
kP
erio
dT
ues-
Thu
rsV
eh.-
hrs
ofD
elay
)30
,000
12,3
0013
,000
40,5
0011
,000
7,90
08,
400
5,30
05,
400
11,6
006,
100
16,7
0022
,700
1,50
0Pe
rcen
tage
of T
otal
PM
Peak
Per
iod
Del
ay70
%76
%48
%46
%39
%41
%82
%54
%47
%19
%74
%37
%66
%83
%